7 open-source security projects spanning SOC engineering, adversarial ML, threat intelligence, and AI-native cybersecurity tooling.
Full home SOC lab — Wazuh + Splunk + Suricata with MITRE ATT&CK mapped detection rules, incident response playbooks, and Active Directory attack simulation.
Production-grade RAG pipeline (~1,500 lines) using LangChain, Claude API, ChromaDB, and Streamlit. Achieves 85% retrieval precision over a cybersecurity knowledge base with semantic search.
BFSI sector threat intelligence — IOC enrichment automation, threat actor profiling, and India-focused threat landscape reports using OTX AlienVault and VirusTotal APIs.
Malware analysis reports, YARA rules, and API security testing templates covering OWASP API Top 10 with Burp Suite findings and remediation guidance.
Machine learning model (Random Forest + SMOTE) to classify SOC alerts as true/false positive — reducing analyst fatigue. Includes Flask API for real-time inference integration.
LLM-powered threat intelligence summariser using Claude API, OTX AlienVault, and RSS feeds to auto-generate daily threat briefs — reducing analyst research time significantly.
FGSM and PGD adversarial attack simulation on neural networks with adversarial training defence. Research-grade PyTorch implementation — BFSI fraud detection context. Based on IDRBT thesis research.